Preliminary data in this Collaborative R01 Project demonstrate the feasibility of screening libraries of small molecules to discover compounds that promote the differentiation of embryonic stem (ES) cells and human pancreatic duct cells. A substantial Institutional investment has been made in chemical libraries, robotics and automated microscopy to perform cell-based screens. The proposed research will build on these successes and this infrastructure by developing high throughput screening technology to discover molecules that promote beta-cell differentiation. A strong synergy is created by the complementary expertise of each laboratory. The Mercola laboratory has led development of cell-based screening efforts at the Institute and provides expertise in robotic screening, ES cell culture and HIV lentiviral technology while the Levine laboratory provides knowledge of diabetes, technology for culture and manipulation of primary human pancreatic duct tissue. Aims 1-3 will develop and perform a pilot screen for molecules that promote ES cells to differentiate into pancreatic endodermal precursors. To do this, we will produce a panel of ES cells with GFP under control of differentiation specific promoters to evaluate differentiation-promoting capability of 10,000 natural product and established drugs. This compound library was chosen for chemical diversity and prior success at yielding modulators of intracellular signaling pathways in several cell types. Aims 4-5 will repeat the screens with adult pancreatic duct cells and an immortalized cell line, both transduced with HIV-lentiviral vector carrying GFP under control of the insulin promoter. An unbiased, chemical approach to beta-cell differentiation is made possible by advances in high throughput technologies, availability of diverse chemical libraries and the ability to easily place reporter proteins such as GFP in relevant cells under the control of genes that characterize distinct stages in pancreatic differentiation. In addition to providing tools to regulate differentiation, knowledge of the "hits" might reveal insight into cellular targets that are involved in pancreatic development, or be used as affinity probes to identify such proteins, in a process that is becoming known as chemical genetics.